Fingerprint Identification Technology of Power IOT Terminal based on Network Traffic Feature

Due to the characteristics of heterogeneous hardware and complex access methods in power Internet of Things (IOT) terminal, it is difficult to achieve fingerprint identification through unified deployment of security probe agents. A fingerprint identification technology for power IOT terminal was proposed based on network traffic feature in order to carry out identity authentication, access control and other security protection of IOT terminal based on traffic fingerprint. According to the characteristics of the hierarchical model of power IOT, the technology extracts a traffic feature matrix by deploying traffic capture, analysis, and fingerprint identification model at the edge of the IOT agents, and then classifies and identifies the extracted traffic fingerprint data based on Multi-class support vector machine (MCSVM) to achieve fingerprint identification for accessing the power IOT terminal. The simulation results show that the technology can complete the classification of common power IOT terminal and identify their device types, and has better fingerprint identification accuracy compared to some machine learning algorithms.